1. A 68.8% of visitors surveyed consider that free Wi-Fi access is an element that generates attractiveness to the Park. People suggest that free public internet access is an interesting amenity that would add value to the public space. Despite this assessment, this does not translate, in the short term, into significant changes in behaviour.   2. Compared to the control condition, business owners' intentions to do certain activities using the internet, rather than analogue forms, increased. Specifically, running the business operations (23.8%, p<0.1) and personal activities -leisure and entertainment such as reading, watching TV series or movies, social networking, streaming, etc.- (35.7%, p<0.1).   3. In an experiment with 8 business units randomly allocated to treatment and control as cluster units, customers who received information about free high-speed internet were 8 times more likely to recommend their family and friends to visit the park, versus those who did not receive this information.   4. Data from the visitor and business owner surveys, as well as results from business observation suggest that perceptions, assessments, and intentions regarding the internet can change, but this does not always lead to behavioural changes in the short term. The effects of this intervention need to be assessed in the longer term.   5. When assessing changes in business behaviour in comparison to control conditions, 2 relevant changes were identified: (i) on the one hand, the use of informal accounting mechanisms increases by 24.7% (p<0.1). This is in line with the contributions of the Digi-Chiquihuites project on inclusive digitalization, where the training programme included the use of 3 digital tools to control expenses and income: Treinta (30), Fintonic and Monefy. (ii) On the other hand, the change in the use of time in activities, a change of 8% (p<0.1) of businesses that make payments for basic services such as water, electricity, telephone, through the internet is observed.   6. Additionally, A/B/n testing on promoting the public space and internet access through social media showed that users interacted more with the A variant, which is the base condition and shows no publicity about the free Wi-Fi. The rest of the 5 variants, underperformed the A variant in the overall interactions indicator. Comments reveal that users tend to associate the Wi-Fi with extra charges to access the public space, although the internet is totally free and the entry to the park remains free as well.   In addition, inputs from a synthetic controls cuasi-experiment are pending to assess the impact of having announced connectivity on the number of visitors to public spaces.

N/A

We had problems with the speed of the internet. Due to weather conditions, one of the access points was damaged and although TIGO, the internet provider, fixed the problem, at some points, the internet speed was not as expected.

Social media data analysis allowed us to analyse the impact of communicating that the park has free and public internet. A geo-referenced assessment of the Wi-Fi network coverage area within the park was carried out, it allowed us to analyse whether the behaviours and perceptions of visitors and business owners changed depending on if they were in the coverage area or not. Extensive literature has been developed around technical and regulatory issues related to the deployment of infrastructure supporting Wi-Fi networks, but there has been limited empirical study of the social implications of these networks (Forlano, 2008; Hampton & Gupta, 2008; Powell, 2008). The RCTs, A/B/n testing, and Diff-in-Diff cuasi experiments carried out contribute to these empirical studies by exploring attitudes and behavioural changes after installing a free Wi-Fi network in public spaces. Specifically, we explored people's motivations for visiting the park, what kind of activities they do in the park, for how long, in the company of whom, and whether they use the internet for some of these activities. With respect to businesses, we inquired about the effect that free public internet access had on their e-formalisation process, on the activities they carry out, and on their intentions to use the internet for certain daily life activities.

A dynamic map was created showing the Wi-Fi access points, the internet coverage area and the behaviours of visitors and tenants with respect to these points. The RCTs and Diff-in-Diff Fixed Effects regressions empirically tested the effect that free public internet access has on people visiting and working in the park. Through a 5 variants social media randomized A/B/n tests we were able to analyse how people react to publications promoting Balboa Park, comparing posts that refer to the free public Wi-Fi access and different uses, such as teleworking, distance education and long distance communication with friends and families. These type of online tests allows to collect large amounts of data in short periods of time and budgets.

- Tigo. Tigo is the commercial brand of the Millicom company and has been present in El Salvador since 1992 and has ranked #1 in the country in mobile technology services. UNDP hired Tigo to provide a 100 Mbps Wi-Fi network in Balboa Park.   - Eurolatina. Eurolatina was hired to carry out the measurements and field observations.

ISTU (Salvadoran Institute of Tourism). ISTU is the entity that manages 3 Natural Parks and 11 Water Parks, including Balboa Park. ISTU provided logistical support and supported the whole process within the Park.

N/A

The interplay of innovation methods and new forms of data is what allowed us to work empirically on this topic at low cost. The implementation was short, so using RCTs allowed us to see the changes, even though it was only a short time after the intervention. The use of social media allowed us to have a large reach, large sample sizes and to get real-time feedback at low cost.